Spring 2015
Industry Study
Final Report
Robotics and Autonomous Systems
The Dwight D. Eisenhower School for National Security and Resource Strategy
National Defense University
Fort McNair, Washington, D.C. 20319-5062
i
ROBOTICS AND AUTONOMOUS SYSTEMS 2015
ABSTRACT: The Robotics and Autonomous Systems (RAS) industry plays a critical role
producing America’s preferred means of global power projection: unmanned aircraft systems. But
in a larger sense, RAS technology has also altered the defense industry’s relationship with DoD,
as the defense sector increasingly finds itself following commercial sector innovations. This report
outlines the major issues affecting the defense sector of the RAS industry and DoD’s ability to
acquire and employ RAS. It then makes recommendations to enable DoD to better capitalize on
the technology, strengthen its relationship with the industry and ultimately improve the nation’s
warfighting capability.
BG Mordechay Baruch, Israeli Defense Force
LTC Clinton Cox, US Army
Mr. Terry Emmert, Office Secretary Defense
COL Daniel Friend, US Army
Mr. Riley Jay, National Geospatial Agency
Lt Col Linell Letendre, US Air Force
Lt Col Robert Masaitis, US Air Force
Mr. David Mico, Dept of State
Lt Col Kevin Murray, US Marine Corps
Lt Col Richard Neitzey, US Marine Corps
Dr. Jeffery Paull, Dept of Navy
CDR Jerome Smith, US Navy
COL Stephanie Tutton, US Army
Mr. Thomas Wilson, Veterans Affairs
Lt Col Lori Winn, US Air Force
CAPT Matthew Pregmon, US Navy, Faculty lead
COL David Shugart, US Army, Faculty
ii
PLACES VISITED
Domestic:
AlphaLab Gear, Pittsburg, PA
Carnegie Mellon University Robotics Institute, Pittsburg, PA
RedZone Robotics, Pittsburg, PA
Astrobotics, Inc., Pittsburg, PA
Re2, Pittsburg, PA
National Robotics Engineering Center, Pittsburg, PA
Human Engineering Research Center, Pittsburg, PA
AAI/Textron Unmanned Systems, Hunt Valley, MD
Federal Aviation Administration, Washington, DC
Office of Naval Research, Ballston, VA
Lockheed Martin Advanced Technology Laboratories, Crystal City, VA
Aurora Flight Sciences, Manassas, VA
Stark Aerospace Corp, Arlington, VA
Naval Air Station Patuxent River, NAVAIR UAS, NAS Pax River, MD
Intuitive Surgical, Sunnyvale, CA
Suitable Technologies, Palo Alto, CA
Institute For The Future, Palo Alto, CA
Stanford Research Institute, Menlo Park, CA
AeroVironment, Simi Valley, CA
Jet Propulsion Lab, Flintridge, CA
Northrop-Grumman, Palmdale, CA
International: Israeli Ministry of Defense, Tel-Aviv, Israel
Roboteam, Tel-Aviv, Israel
Cogniteam, Tel-Aviv, Israel
G-Nius Unmanned Ground Systems, Nazareth, Israel
Rafael Advanced Defense Systems Ltd., Haifa, Israel
Amstaf on Guard/Automotive Robotic Industry Ltd., Nazareth, Israel
US Embassy, Tel-Aviv, Israel
Ben-Gurion University of Negev, Autonomous Robotics Lab, Beer Sheva, Israel
Aeronautics Defense Systems & Unmanned Systems Ltd., Yavne, Israel
SimLat Unmanned Vehicle Systems, Herzliya, Israel
Israel Aerospace Industry, Tel-Aviv, Israel
1
In addition to new technologies, a third offset strategy will require innovative
thinking, the development of new operational concepts, new ways of organizing,
and long-term strategies. . . . [Y]ou need to ask how should we prepare for a future
where new and disruptive technological developments are continuously occurring?1
—The Honorable Mr. Robert Work
Deputy Secretary of Defense
Destiny is no matter of chance. It is a matter of choice. It is not a thing to be waited
for, it is a thing to be achieved.2
—William Jennings Bryan, American’s Mission
INTRODUCTION
Over the course of the past fifty years, robots have sprung forth from the realm of science
fiction to become essential, if still unintelligent, coworkers and comrades for industries and
militaries worldwide. But the past fifty years will pale in comparison to what the next 50 hold in
store not for just commercial and military endeavors, but for society and humanity. The confluence
of revolutionary advancements in computing power, multispectral sensors, miniaturized inertial
positioning systems, and lightweight materials is already altering the landscape of human
technological achievement. Witness the rise of hardware/software integration giants like Google
and Apple, the likes of which have greatly incentivized the application of software engineering to
monetizable problems. Within this primordial mix of technology, societal demand and financial
incentive, exist the building blocks of a true revolution in the application of robotics and
autonomous systems (RAS) to the human endeavor.
While the RAS industry itself remains diffuse and vast, its impact on American society will
be felt along two critical dimensions. First, the cost of robotic systems will continue to fall to the
point where American industry, including manufacturing, may see a resurgence based on the use
of highly automated robotic factories. While a potential boon for American firms seeking to
“reshore” their previously offshored manufacturing activities, robotics will enable far more than
just the resurgence of America’s manufacturing sector. Robotic and autonomy technologies will
usurp many jobs currently performed by unskilled, semi-skilled, and in some cases skilled laborers.
The impacts on the US job market could be severe as thousands or millions of workers are
gradually displaced and the very job market itself is redefined by new skills required to supervise
robotic and autonomous systems. For the average person, the rapidly modernizing world is
effectively getting easier and more difficult to live in simultaneously. (See Essay 1.)
The other dimension of the RAS industry is the development and employment of military
power to ensure the nation’s security. Since prehistoric men first picked up stones to hurl at their
foes, competitive advantage in war has gone to the group most capable of using new technology.3
The security environment of the 21st century is not so different. While the tools of war have
changed drastically, the principle of using technology for efficiency in the exercise of national
power remains essential. With funding for defense resources diminishing for the foreseeable
future, the U.S. will need to develop increasingly efficient means by which to project force. The
nascent RAS industry holds a key not only to greater efficiencies in force projection, but also to
“offsetting” America's potential military rivals. The adoption of RAS is not simply the next
evolutionary technological step, nor just the next trendy idea; it has the potential to support the
transformation of how the U.S. achieves its national security objectives.
2
Forging the nation’s third “offset” strategy depends not upon matching symmetrical force
capabilities or increasing Department of Defense (DoD) resource levels to police the globe.
Instead, the US military’s ability to counter the conglomerate of disparate rivals rests—as it has
since the end of WWII—on the abilities of its industrial base to deliver technological warfighting
advantages. Even though America has achieved technological military superiority, the evolution
of military power is dynamic, and its rivals continue to counter those advantages with tactics and
technologies of their own. Thus, American’s ability to deter, dissuade and otherwise influence
nations and non-state actors in far-flung regions requires, above all else, the ability to leverage
industry’s ability to continuously innovate. This seminar focused its study on the capability of the RAS industry to meet US national
security objectives. To evaluate this industrial capability, we conducted academic industry
research, field studies, interviews with subject matter experts, and focused analysis of the US
unmanned systems defense sector. We began by engaging with recognized thinkers in robotic
research and reviewing leading think tank reports and DoD’s RAS-related plans in order to map
the industry’s conceptual space. We compared and contrasted strengths, weaknesses and linkages
among academia, publicly funded research centers, and private industry—from small start-ups in
the commercial sector to large defense corporations. To evaluate the US industry’s status against
the global market, we traveled throughout Israel and viewed an alternative model for employing
RAS technology and leveraging an industrial base in pursuit of a nation’s security objectives. We assess the US military is acquiring a disparate portfolio of RAS-related systems with
no unifying RAS vision and, consequently, the defense industry is coping with a high level of
uncertainty. More importantly, however, the U.S. lacks a focus toward maximizing the use of
autonomous systems to increase the nation’s ability to project power. In the worst case scenario,
the U.S. could find its current force structure increasingly contested by remotely operated and
unmanned systems fielded by an adversary ready to capitalize on such opportunities. These
challenges stem not from technological difficulty—although much remains there—but a lack of
forward-looking doctrinal development. As a result, industry is uncertain about DoD’s level of
commitment to adopting RAS and unsure about what types of RAS investment will best advance
national security interests.4
To support this assessment, this report first defines and assesses the defense sector of the
RAS industry. Next, we explore the structural and institutional challenges facing the industry and
the US military’s efforts to adopt this technology. Finally, we provide a glimpse of a potential
RAS innovation ecosystem that could bolster the nation’s ability to quickly and efficiently harness
cutting-edge RAS for the purpose of national defense. We conclude by offering a series of
recommendations to achieve that vision.
3
INDUSTRY DEFINED
Just as the concept of “what is a robot”
is broad and varied, so too is the loosely-
defined RAS industry. The RAS market can
be thought of as the group of firms that
produce everything from unmanned military
systems to intelligent software for use in
“learning” capable industrial machines. (See
Figure 1.) Its products are the synthesis of
multiple engineering disciplines, including
software, electronic and mechanical. The
rate of advancement in the RAS industry is
inextricably linked to advances in several
related technologies, to include
microprocessors, batteries, mechanical effectors, sensors, and composite materials.
In the commercial sector, the majority of robotic systems are used in industrial
manufacturing, though the past few years have seen a rapid expansion in robotic applications, to
include entry into the medical device, food processing, and electronics manufacturing spaces. US
manufacturing firms now use an estimated 230,000 robots, second only to Japan with 300,000
industrial robots.5 Other segments are slowly taking shape and are becoming increasingly lucrative
enterprises. The household and consumer segment aims to transform everyday lives through the
application of autonomy technology applied to home use such as vacuum cleaners, lawn mowers,
pool cleaners, personal assistants, etc.6
The defense sector is more tightly focused on the development and manufacturing of
unmanned or remotely operated vehicles of varying levels of autonomy that operate in the air,
land, sea, and undersea domains. This sector is characterized by a mix of large, established defense
industrial firms, as well as a smattering of small-to-medium sized firms that entered the market by
fielding small robotic systems developed for use in Afghanistan and Iraq.
As RAS technology is a relatively recent addition to military operations, the industry’s
lifecycle dynamics are best explained through the concepts of disruptive and sustaining
technologies.7 (See Figure 2.)8
Disruptive technologies tend to
upset the established order of an
organization when introduced.
For example, the unique
capability of the RQ-1 Predator
to provide senior commanders
full-motion video in real-time
changed the nature of battlefield
command and control when it
was fielded in the mid-1990s.
Sustaining technologies, on the other hand, tend to fit in the shakeout or production portion of the
curve. The follow-on to the Predator system, General Atomics’ MQ-9 Reaper, is an example of a
sustaining system and is essentially a bigger, better version of the Predator.
Conversely, doctrinally upsetting systems like the Navy’s UCLASS (Unmanned Carrier-
Launched Airborne Surveillance and Strike) prototypes tend to remain in the entry and
4
experimentation phase while the sponsoring service determines what capabilities the system
should ultimately possess. As disruptive capabilities proliferate and become the industry norm,
firms are faced with the strategic dilemma of either refining previous products or developing new
market segments through product differentiation. This tension has become the defining
characteristic of the defense unmanned systems market space.
CURRENT INDUSTRY CONDITIONS
Industry Segment Analysis. Due to the immaturity of the majority of the national security
RAS industry, our analysis of defense industrial firms focuses primarily (but not exclusively) on
those competing in the unmanned air vehicle (UAV) sector. UAVs represent the most mature
market and provide insight into likely future trends as the DoD fields larger numbers of unmanned
systems in the maritime and ground domains.9 While some public financial data exists on the UAV
sector, analysis is challenging as the UAV manufacturing sector possesses no North American
Industry Classification System identifier and few large defense companies distinguish these
systems within their military sales figures. The large number of competitors within the UAV
segment presents the best data from which to assess its likely future. Existing since the late 1970s,
the maturation process of the UAV segment is instructive for divining future trends for other
unmanned system segments.
Within the UAV segment, DoD categorizes systems by weight and operating altitude,
known as groups 1-5.10 (See Figure 3.) Each category poses distinct challenges in terms of
complexity, technology, payload, capability, and price differences.
The past decade has been marked by a relative oligopoly in each
of the group sizes. AeroVironment and Boeing’s Insitu dominated
Group 1 and 2 (small UAS or SUAS), respectively. AAI
Corporation’s Hunter and Shadow platforms controlled the Group
3 market. General Atomics’ Predator and Reaper systems were
almost exclusive in the Medium Altitude, Long Endurance
(MALE) Group 4 into Group 5 category. Finally, Northrop
Grumman’s Global Hawk system makes up the High Altitude,
Long Endurance (HALE) Group 5 market.
In sum, the UAV segment is characterized by one to two companies dominating a given
group market and facing limited competition as a reward for being the first to market to provide
intelligence, surveillance, and reconnaissance (ISR) solutions for combat operations. With two
simultaneous long-term combat operations (Enduring Freedom and Iraqi Freedom), the
proliferation of Overseas Contingency Operations (OCO) funding allowed quick fielding of these
systems and created above average returns on investment (ROI) with an 11.5% growth in revenues
from 2007 to 2012 for the few companies involved across the defense industry.11 Recently, the
federal government’s reduction in purchases, combined with delays in expanding into the
commercial market, has altered the structure and strategy of the UAV firms as detailed below.
Market Structure (Porter’s Five Forces Model). Porter’s Five Forces model
demonstrates the highly competitive nature of the RAS market and explains why firms in this
industry seek to create new markets through niche strategies to secure a competitive advantage.12
5
Bargaining Power of Buyer: High. The most significant factor facing the industry is the
buyer’s bargaining power. (See Figure 4.) The
US government has a relative monopsony on
the UAV market with 90% of the sales
revenue.13 The monopsony impact is magnified
by: 1) reduction in market sales due to
decreased DoD budgets, 2) inhibition of UAV
commercial market by Federal Aviation
Administration (FAA) delays with opening the
National Air Space (NAS) to UAVs and 3)
highly controlled international sales under the
International Traffic in Arms Regulations
(ITAR) and the Missile Technology Control
Regime (MTCR).
Rivalry Among Competitors: Medium. In 2012, the concentration ratio was 59.1% for the
industry overall.14 The competition between rivals still resembles an oligopoly in the group 3, 4,
and 5 categories as a few large companies vie for a limited number of large dollar US government
procurements. In groups 1 and 2, the expansion of competitors for each program makes the market
more reflective of monopolistic competition.
Threat of New Entrant: Low. The challenges of government acquisition, delayed opening
of the commercial UAV market and the costs of remaining technologically ahead of the field
combine to keep the threat of new entrants low.
Threat of Substitutes: Medium. The US military’s advocacy for the primacy of manned
systems over unmanned drives the availability of substitutes. Additionally, the supporting
architecture (e.g., data-link satellites) that enables long-distance remote UAV operations is
constrained. This limitation has created a substitute marketplace for both leased and purchased
manned ISR capabilities.
Bargaining Power of Supplier: Low. The increased obtainability of miniaturized
technologies needed for UAV production renders the supplier’s bargaining power low. This
bargaining power may alter in the coming years, however, as large civilian corporations purchase
small firms with emerging UAV component technologies. Additionally, the advancement of open
architecture will allow suppliers with sensor, communication, or manipulator systems to contract
directly with the government thus increasing their power as the platform commoditizes.
Firms’ Conduct (Strategic Gameboard Analysis). Though intended for individual firm
analysis, the lens of the strategic
gameboard provides insight into where
the majority of firms are targeting their
strategy.15 (See Figure 5.) In support of
combat operations over the last decade,
most UAV companies created new
markets with their products as the
military quickly fielded this new
technology. Once the immediate
requirement was met and sales leveled
off, companies sought to further segment
the market and create niche capabilities
6
through focused research and development (R&D). During the OCO funding years, profit margins
and gross sales created value and supported these competitive strategies. The reductions in defense
funding and limitations on OCO have changed the market strategy across the industry.16 Most
companies are seeking new markets in the commercial and international markets arena and life
cycle support of existing products to create long-term profits.17 This may be the only viable
strategy for the smaller companies that are largely dependent on UAV sales. Larger defense
companies appear to be weathering this fiscally constrained storm through reduced R&D, cost-
cutting moves, and buyback of stocks with excess cash reserves to create necessary dividends for
their stockholders.
Industry Performance. The domestic UAV industry has demonstrated a slow-down in
revenue growth that mirrors the reduction in defense spending over the last four years. The overall
spending within the UAV industry has declined 6.2% annually from 2010 to 2015.18 Despite this
recent defense spending recession, The Teal Group continues to estimate that worldwide UAV
industry sales will grow from $6.4 billion to $11.5 billion in the next ten years.19 Their forecast is
based on growth in the commercial and international sectors, which are delayed. The small
companies in the UAV market are producing an ROI well below the ten percent value accepted as
the cutoff for creating value. For example, since the market decline, AAI Textron reported a 5.4%
ROI in 2013 while AeroVironment had a 5% ROI in 2014.
In order to create value going forward, these companies must continue to support the R&D
required to capture new programs. In addition, they must adjust their strategy to diversify their
market to international sales or capture the opening of the commercial market. The large defense
contractors continue to deliver strong ROI through their diverse portfolios. Over the past few years,
Northrop Grumman had an ROI of 12.4% and Lockheed Martin had an ROI of 29.8% across their
range of business sectors. These large firms’ continued interest in the UAV industry depends upon
a sufficient market existing to justify their R&D investment. The limited number of DoD programs
of record (POR) will result in heated competition for the decreased availability of defense market
share. The required commitment in independent research and development (IR&D) funding to
compete for large programs, like the UCLASS, will have negative consequences for those
companies not selected. Potential growth areas for profitability within the UAV market are fee-
for-service and performance-based logistics (PBL) contracts that could provide additional revenue
streams to the contractors above profits gained from system sales.
International Market Forces. The global market for UAV is extremely competitive as
foreign companies increase their presence in this sector.20 Israel and China lead these efforts with
aggressive sales of multiple platforms that exceed or have near equivalent capabilities to available
US defense products. Without public data, it is almost impossible to determine the exact market
share that they command; however, it is clear that these international competitors have surpassed
US defense industry sales. For the years 2005-2012, Israel exported $4.6 billion in UAVs—close
to twice that of the U.S.21 Israel’s compelling innovation model and export emphasis provides
strong evidence for their continued market strength. (See Essay 2.) The persistent challenges of
ITAR and MTCR limit US companies from fully realizing the benefits from overseas sales and
the support that these sales would bring to the domestic defense industrial base.
Outside Influences on the UAV Market. Outside influencers and stakeholders continue
to shape the UAV domestic industrial base. Congress has sought to support the industry through
legislation directing the DoD to detail specific actions with regards to the procurement strategy.
Congress has also served as the forcing function to spur the FAA to shorten timelines to open the
NAS for commercial UAV use. Finally, trade groups in support of the industry expansion have
7
lobbied Congress to put pressure on DoD and DoS to change policies currently impacting domestic
and overseas sales of UAV.
CHALLENGES: INDUSTRY AND DOD OBSTACLES LIMITING POTENTIAL
BENEFITS OF RAS
The nascent RAS industry faces significant challenges in gaining both acceptance and
adoption into the US military’s way of war, though the reluctance of the US military to accept new
technologies is not a new phenomenon. Militaries have historically rejected groundbreaking
technology. Such new technologies are best described as disruptive innovations—improvements
in performance “along a war fighting trajectory that traditionally has not been valued.”22 The
difficulty in understanding the particular barriers to adoption of robotic and autonomous systems
is compounded by the diffuse nature of the technology. While the services are procuring individual
weapons systems that can be variously described as either “robotic” and/or “autonomous,” the
underlying technology—increasingly powerful microprocessors, miniaturized inertial systems and
sensors—is increasingly employed on weapons systems not ordinarily considered as either robotic
or autonomous.23 With technology advancing along multiple simultaneous dimensions, the central
challenge is not the identification of promising new technology, but rather divining future
warfighting concepts that can capitalize on that technology. Without a deliberate and iterative
effort to develop such concepts, industry will continue to receive mixed demand signals and guess
about where to focus its R&D efforts.
Challenge 1: Lack of Coherent, Integrated Warfighting Vision. Autonomy technology
holds the potential to fundamentally alter the US’s military advantage over its future adversaries.
However, the individual military services’ approach to RAS reflects the historical pattern of
procuring and prioritizing weapons systems that advance well-established means of warfighting,
but are also heavily anchored in the past decade of conflict. Consequently, few forward-looking
concepts of operations (CONOPS) exist for exploiting RAS’s potential advantages in the context
of joint warfighting. Instead of becoming a force-wide attribute, autonomy (or the degree thereof)
has become a means to accomplish undesirable missions.24,25
Ultimately, individual military services decide which systems and capabilities align with
their core interests, resulting in the funding of a disparate and parochial set of capabilities. As such,
the services’ requirements for future unmanned systems tend to shift frequently as debates about
roles and missions occur within the services.26 Industry’s response has been to take a conservative
approach toward developing new concepts while DoD figures out what roles its future unmanned
platforms will perform. Combined with DoD’s declining R&D budget, even the larger defense
conglomerates are unsure to what degree the DoD is willing to invest in RAS.27
Challenge 2: Culture and the Disutility of the “3D” Paradigm. The services also face
internal challenges in terms of the degree of autonomy disrupting their established systems of
domain–centric warfare. The predominant attitude is one of grudging acceptance when it has come
time to replace humans with mechanical and robotic systems, with three notable exceptions:
missions and tasks that are either too dangerous, too dirty, or ultimately too dull for humans to
do.28 This “3D” mentality has long been the mantra guiding the adoption of RAS technology, but
limits critical thought about its applicability to DoD opportunities. Furthermore, some military
communities (i.e., pilots, ship drivers, logisticians, etc.) view RAS systems as a direct threat to
their core skill sets and tasks—not to mention manpower authorizations—that form the very
identity of their military services.
8
The result is that within the services (particular the aviation-oriented communities)
unmanned systems duty is a lesser professional occupation than operating manned platforms.
Despite the fact unmanned systems have been in use for decades, their operators have not risen to
the highest ranks of their respective services. Without senior officers who understand the unique
opportunities unmanned and remotely operated systems offer, few requirements for new,
revolutionary systems are likely to survive against the entrenched way of conducting business.
Challenge 3: Lack of Trust in RAS Technology. The cutting-edge RAS technology this
group observed, while impressive, still trails by a wide margin the cultural perception of what
robotics ought to be able to do. The fragile and essentially disposable nature of the first generation
of RAS also contributes to a lack of trust in the overall technology, even though early systems
generally performed as designed. As a result, the services are hesitant to build future concepts of
operation that rely on unproven and yet-to-be-developed military for mission sets outside the “3D”
paradigm.
The task of identifying and developing promising RAS technology falls primarily to the
individual services’ research labs and the Defense Advanced Research Projects Agency
(DARPA).29 Many times the successful adoption of technology hinges on its inherent
demonstrability. RAS demonstrations tend to include large numbers of operators and support,
which gives the impression of little to no savings in manpower cost. Demonstrations also often
make public exhibitions of system failures and limited successes, which can lead to lack of
confidence in not just the individual system, but the state of the underlying technology.30
Challenge 4: The Defense Acquisition System. Failures of the acquisition system to
deliver advanced technological programs on time, on budget and performing up to specifications
are not new. The timeline from idea and requirements development to fielding a fully operational
system almost guarantees that the acquired system will have been technologically surpassed by the
time it reaches a warfighting unit. This elongated process does not mean that the technology won’t
be useful, but it does pose problems. Identifying future state-of-the-art RAS capabilities is difficult;
codifying them into meaningful performance parameters is an exercise in hedging for uncertainty.
Challenge 5: Budgetary Uncertainty. One of the most significant challenges facing the
near-term development and fielding of RAS are the individual service’s existing weapons system
programs. The uncertainty as to future funding levels has caused the services to rally in defense of
large spending programs that sustain their most cherished platforms. This squeezing effect limits
the services’ ability to invest in future programs that require the development of unproven and
immature technology, but could ultimately address key security challenges. Additionally, the ad
hoc nature and rapid acquisitions authority under which some of the urgent wartime systems
(primarily the thousands of explosive ordinance devise (EOD) robots currently fielded in
Afghanistan) were procured means there is no long-term POR, and therefore no sustainment
funding for future use and improvement.31 Corralling these systems into a consolidated program
now competes directly with funding for other, more forward-looking service programs.
Challenge 6: Private Sector Competition. As a corollary to the defense industry, the
commercial sector increasingly rewards the development of software as opposed to robotic
hardware. As a result, commercial firms offer higher-paying positions to professionals in the
robotics field and are viewed as more attractive than defense industry positions.32 Google, for
example, employs at least one former DARPA director and has recently purchased eight robotics
firms, two of which are DARPA Robotics Challenge winners.33
The software and applications market also represents a far more lucrative endeavor for
individual firms, particularly when viewed against the current defense budget environment. Firms
9
like Google, Amazon, and Apple are not currently investing R&D in military-ready technologies.
More importantly, the potential profits from DoD projects do not come close to the profits such
firms can earn in the commercial market. The private sector—once a follower of DoD
technology—has become a market-driven leader in many RAS-related technologies.
OUTLOOK: CHARACTERISTICS OF AN IMPROVED RAS INNOVATION
ECOSYSTEM
When considering the future outlook of RAS outside of the defense context, it is clear that
the civilian sector will see explosive growth in RAS as the technology matures and becomes
cheaper. With the benefit of a profit motive, the commercial sector will continue to outpace the
defense industry in researching and developing RAS applications. The commercial RAS sector
will make these technological leaps out of sight of DoD, as the norm for the larger innovating
firms is to keep products wrapped in secrecy until ready for launch. In light of this unfamiliar
dynamic, DoD needs to adjust its view of how to work with commercial RAS firms and learn to
leverage the inherent creativity and speed with which the commercial sector can match
technological solutions to seemingly unsolvable problem sets.
Given this outlook, this seminar sought to better understand and define an ecosystem with
the characteristics necessary to promote rapid RAS innovation within the defense industrial sector.
To further our understanding, we visited and studied both domestic and international innovation
models and gained a new perspective for an innovation ecosystem that could better harness and
integrate RAS technology. This section outlines the lessons we took from these engagements and
describes the key attributes of a system that will allow DoD to not only move past its current
challenges, but to take the lead in developing military robotics and fielding useful capabilities in a
timely fashion.
During our field studies to Pittsburg, Pennsylvania, we saw first-hand the use of a
commercial “accelerator” model. These firms provide startup companies with training, mentorship
and office space as well as a quick infusion of seed funding; in exchange, accelerators receive a
small share of equity in these companies.34 Besides providing a quick boost of capital and
mentoring to a cohort of startup companies, the networks of learning within accelerators provide
the keystone of the innovation model.35 Accelerators typically surge development on a fixed
timeline, taking products from concept to product in a matter of months.
We also examined an alternative paradigm for interactions between the military and
defense industrial base through extensive field studies in Israel. The close working relationship
between Israel’s Ministry of Defense (MoD), Israeli Defense Forces (IDF) and its defense
industrial base provided a unique perspective on harnessing technology quickly to solve real-world
battlefield problems. While Israel’s strategic focus and geographic position are drastically different
than that of the U.S., some of the lessons in promoting innovation in the RAS industry are highly
applicable. (See Essay 2.)
Based on our studies, this seminar believes the nature of DoD and RAS industry
interactions ought to possess four key characteristics. First, the development and idea generation
process needs to feature a close working relationship between the developers and the end users.
This relationship is symbiotic in that the developer gains insight into potential military utility while
the operator gains a reciprocal understanding of the art of the possible. Operators and developers
working together provide a greater chance of maturing an idea that can be written into both a useful
operational concept and a realistic set of requirements. Second, DoD needs a bridging function to
shepherd promising systems from the idea stage through the requirements-generation stage. Third,
10
the ecosystem must emphasize demonstrations of RAS technology to the services’ senior
leadership. Demonstrations for senior civilians, congress, and others are useful, but historically
none of those are as effective as gaining the advocacy of a service’s chief for a particular idea.
Lastly, the ecosystem must feature more flexible funding mechanisms that will enable rapid
technology procurement and encourage small firm participation in the defense sector of the RAS
ecosystem.
RECOMMENDATIONS
To overcome these challenges in the RAS industry and ensure that the U.S. obtains the full
value of RAS contributions to national security, both DoD and industry must improve how they
think about, plan for, develop, and utilize robotic and autonomous systems. Based upon our study,
we offer the following recommendations to further that effort:
Recommendation 1: Develop and Manage an Evolutionary RAS Strategy. First and
foremost, DoD must intentionally develop and manage an evolutionary strategy toward RAS in
order to lead to revolutionary results. While this seminar is convinced that autonomous military
systems will one day lead to a revolution in military affairs, such results will not be achieved in
the next five to ten years. To encourage progress toward such an outcome, DoD should target areas
where RAS integration into the military will not collide with institutional resistance and focus on
uses of RAS that will build trust and acceptance of the end-users and military leadership. Using
the advice from Terry Pierce, author of Warfighting and Disruptive Technologies, RAS product
champions should “disguise or shape the disruptive innovation as a sustaining innovation” to
improve the likelihood of gaining acceptance within DoD.36 To that end, DoD should focus upon:
Logistics: RAS has great potential to increase logistics productivity in everything from
warehouse management to resupply routes. The logistics arena also allows for quick, easy wins by
adopting robotic and autonomous systems that have been proven to work in the commercial sector
(e.g., Amazon robot-assisted warehousing). An iterative approach to integrating RAS into logistics
is required to gain user confidence and promote innovation gains. (See Essay 3.)
Manned-Unmanned Teaming (MUM-T): DoD should develop CONOPs and then seek to
procure systems that focus on collaboration and synergy between manned and unmanned systems.
Instead of perfecting a robotic system that can “do it all”, services should focus on a “system-of-
systems” approach that enables humans and robots to perform functions for which they are best
designed. Systems such as leader-follower convoys or unmanned-wingman concepts will provide
effects unobtainable by current systems and such an approach will also improve trust and
acceptance of RAS within the defense community. (See Essay 4.)
Leverage commercial RAS Technology: DoD can speed implementation of fielding
systems by leveraged commercial capabilities with a high technology readiness level. One clear
area to capitalize on synergies between the military and commercial sectors involves autonomous
cars and corresponding application to military convoys.
Recommendation 2: Complete the Joint Unifying Vision for RAS. To guide this
evolutionary path, DoD senior leaders must commit to and promote the possibilities afforded by
RAS. To that end, we encourage the Joint Staff to complete and publish a comprehensive and
unifying vision of RAS that will provide direction and focus areas of joint development for the
services across all domains. Additionally, service leaders must do more than just tout their
approval of unmanned systems through notional roadmaps; they must demonstrate that support by
drafting CONOPs that identify how RAS can advance each warfighting domain, develop and
promotes personnel expertise to utilize RAS systems, and budget funds in the POM (Program
11
Objective Memorandum) to create actual RAS programs of record vice reliance on OCO
expenditures.
Recommendation 3: Establish a RAS-specific Board Under the DoD Innovation
Initiative. To underscore commitment to and obtain the most value out of robotics, the Deputy
Secretary of Defense should chair a RAS-specific board under the new DoD Innovation
Initiative.37 This board would be chartered to capitalize on RAS technology identified in the Long-
Range Research and Development Program Plan due to be completed in summer 2015.38 This
board would also be charged to implement a common vision across services; find, sponsor, and
execute testing of early stage technology in an operational environment; and synchronize efforts
and collaboration across warfighting domains. The creation of such a board also enables flag-level
champions to emerge from the services and to promote RAS resource allocation.
Recommendation 4: Adopt a Fast-Follower Approach. Though countless critics have
bemoaned the lengthy procurement process for traditional systems, a protracted timeline for RAS
undermines—and potentially undercuts—the very value of obtaining a high-tech, innovative
robotic systems. While we witnessed amazing technological advances and ideas at the concept or
demonstrator level, moving those ideas across the “valley of death” to development and
procurement has proved challenging. Based on fundamental differences in business models and
motivations, the DIB is simply not structured to innovate with cycle times comparable to the
commercial RAS technology sector. In order to best capture promising RAS commercial
technology for small scale application and develop paths to a traditional programs of record, the
US government can enable the DIB to be a more effective fast follower:
Understand Distinctions between DIB and Commercial Sector: The President should direct
an industry study under his Title III Defense Production Act authority to understand the robotic
industry capabilities, assess any personnel or talent shortages, and clarify which companies are
capable of developing autonomous systems for future defense use. In particular, this study should
determine how the defense industrial base can leverage innovations within the commercial robotics
sector and target R&D allocation based on determined gaps between civilian and defense priorities.
Based on this study, the U.S. should then promulgate a national RAS strategy that encompasses
and promotes development of economic growth and defense strategy resourcing. (See Essay 5.)
Focus R&D that Leverages Commercial Sector Advances: With the decline of research
dollars in the traditional DIB and the growth of powerful commercial companies like Google
overtaking new tech areas like autonomy, DoD must critically examine how it fosters R&D in
RAS. DoD’s current ambiguous commitment to RAS does little to encourage industry innovation
and expenditure of IR&D by the defense firms.39 As one industry representative noted, “Why
should we invest our dollars when you don’t know what you [DoD] want?” A better articulation
of RAS requirements and DoD RAS vision will provide industry assurance and direction for
expenditure of IR&D on autonomous systems. This is especially critical in the “valley of death”
period between science and technology R&D and pre-program of record R&D. Simply relying on
the DIB to provide innovation in this field, however, is not enough; DoD must improve its ability
to leverage commercial markets. By identifying those areas that are of shared interest between the
commercial and military sectors, DoD can prioritize government R&D funds appropriately to
either stimulate a joint effort to develop a RAS capability faster, or to efficiently assign funds in
areas specific relative to DoD interests, leaving other areas that are commercially viable to internal
R&D funding without augmentation. (See Essay 5.)
Create opportunities for DIB and Commercial Collaboration: DoD should adopt a variety
of commercial innovation models such as accelerators, technology sprints, “maker” spaces, and
12
crowd-sourcing initiatives. The commercial sector has found these approaches result in more rapid
generation of ideas, reduction of risk, and greater speed of technology to market. Additionally,
DoD, in consultation with industry, should identify areas where standardization can occur. For
example, the UGV community identified interoperability—enabled through video, audio, and
electrical communications—as a standard.40 Similarly, the cyber community realized the benefits
of standardization through the Joint Open Architecture Spectrum Infrastructure effort by bringing
together industry experts on communication and electronic warfare systems and RF spectrum
standards.41 The key component for RAS will not be the platform technology itself, but the ability
for industry to develop new “apps” and plug-and-play systems to augment the platform. Thus, the
more DoD embraces standardization and common control architecture the greater flexibility it will
have to decouple hardware from software on autonomous systems.
Incorporate Israel’s Building-Block Approach: Israel views unmanned systems not as a
UAV or USV but as a UXV.42 As one Israeli robotics researcher explained, “Only 5-10% of an
unmanned system is unique to a particular domain…so a generic approach is best.”43 This
building-block approach enables more rapid development and eases the test burden by relying on
already validated sub-systems. Adoption a UXV methodology in both procurement and testing
would allow DoD to acquire upgraded RAS technology rapidly and promote interoperability
across the services. (See Essay 2.)
Create Opportunities for Experimentation: DoD should develop a National Autonomous
Test Range that includes robust modeling and simulation tools RAS and physical range space for
MUM-T test and evaluation (T&E). Creating a space that both the DIB and commercial sectors
can utilize would encourage additional collaboration between the sectors, speed technical
improvements in RAS, advance technical understanding of autonomy through a centralized testing
databank, and reduce the overall cost of T&E for RAS. (See Essay 6.) DoD should also require
RAS participation in joint force and service-specific exercises. A combination of operational
experimentation and realistic modeling and simulation tools will have the added benefit of
promoting user trust of RAS.
Recommendation 5: Development of Trust in RAS Users. As with any nascent
technology, industry must avoid overselling technological capabilities—especially early in the
development phase in front of potential operators. Until users begin to trust RAS capabilities,
premature integration or overinflated expectations can actually set back operational acceptance.44
To that end, DoD and industry should replace the “dull, dirty, and dangerous” mantra long used to
describe traditional robotics applications with a new phrase: “efficient, effective, and economical.”
By emphasizing how robotics adds to DoD capabilities vice replaces personnel or missions,
industry can promote a wider array of RAS capabilities and potentials for the warfighter. Finally,
DoD should also look for opportunities to promote talent and robotics capabilities within its ranks
and develop a core of professionals whose in-depth understanding of autonomy naturally leads to
greater trust in the systems. One such way to imbed such talent in through integration of robotics
into the military service academies’ curricula, to include specific concentrations in the study of
robotics. The United States has only begun to start down this evolutionary path, and it needs talent
to propel and guide it toward a robotics revolution.
CONCLUSIONS
To be clear, the U.S. has not lost the robotics revolution. The RAS industry is expanding,
and set to truly transform the nature of society’s use of technology. The DoD remains
technologically ahead of its competitors in many respects. In truth, DoD—largely through the
13
innovative work of DARPA—has done more for commercial robotics around the world than
perhaps any other single entity. But the days of DoD leading industry and dictating terms to a wide
field of capable, defense-focused firms are over. Today’s environment is one characterized by
widespread diffusion of RAS’s underlying technology and the lucrative prospect of its commercial
utility. Just as with GPS and the Internet, many successful robotics firms owe their good fortunes
in some capacity to a previous DARPA-funded research initiative.
For all the collective technical brilliance located there, Silicon Valley, Boston, and
Pittsburgh—collectively America’s RAS center of gravity—will not of their own accord develop
militarily-useful technologies, nor seek to do large-scale business with DoD. For its part, the DoD
may not entirely drive technological innovation within the industry, but it must be capable of
understanding the state of technology for the purpose of assessing its potential defense
applications. To that end, warfighters must be not only capable of developing visionary ways of
warfare, but also articulating the performance attributes of the RAS that will enable them. Until
that happens, the defense industry will continue to be both fractious and focused at the same time:
fractious as it searches for diverse product portfolios with which to assure income, and focused on
incremental improvements to the systems it already produces. The defense industry will persist in
this condition until DoD and the services come together on a collective vision and organizing
construct.
APPENDIX: ESSAYS
Essay 1: Commercial Robotics and the US Economy…………………………………..14
By Lt Col Rob Masaitis, USAF
Essay 2: Israeli Leadership in Robotics and Autonomous Systems…………………….16
By BG Mordechay Baruch, Israeli Defense Force and
Mr. David Mico, Department of State
Essay 3: RAS Integration into Logistics………………………………………………...17
By Mr. Terence Emmert, Department of Defense
Essay 4: Opportunities for Manned-Unmanned Teaming (MUM-T)………………...…18
By LTC Kevin Murray, USMC
Essay 5: DIB, Google, and the Need to Strengthen the Defense Production Act……….19
By Lt Col Linell Letendre, USAF
Essay 6: Recommendations to Improve RAS Testing…………………………………..21
By Lt Col Linell Letendre, USAF
14
ESSAY 1: COMMERCIAL ROBOTICS AND THE US ECONOMY
Robotic and autonomous systems hold the potential to revolutionize American
productivity, create new product markets, and ultimately boost US economic growth—but only if
Americans are willing to go “all in” on the technology. In order for this benefit to materialize,
American firms will have to make difficult investment decisions about technology that is sure to
displace large portions of their human workforces. Additionally, the nation will need to adopt
policies that transform it into the world’s leading adopter of robotic technology.
The stakes for America’s economy are high. Post-2008 financial crisis growth has been
lackluster, and after multiple rounds of debt relief and taxpayer-funded “quantitative easing,” the
U.S. is more than $13 trillion dollars in debt.45 By 2024, servicing just the interest on that debt
will require 3.3% of US GDP, meaning the U.S. will spend more on interest than the nation’s
defense.46 Put simply, the nation’s ability to provide all the things expected of it far exceeds the
US historical 2% GDP growth average. The central question, however, remains: Will the adoption
of commercial robotics be sufficient to lift the US economy out of its stunted growth condition, or
is the idea of a “reshoring” revolution just a hopeful idea?
The rapidly decreasing costs of industrial robots have provided an increasingly suitable
substitute for expensive US human labor, sparking optimism in what has become known variously
as the “reshoring,” “onshoring,” or “insourcing” movement. Bringing back previously outsourced
manufacturing capacity is a promising trend for the economic outlook.47 Having steadily declined
from its peak in 1979, the US manufacturing sector lost more than a 5.8 million jobs (33.6%)
between 2000-2009.48 While a reshoring movement marks the bottom of that curve, the degree to
which more American firms will reshore and its ultimate effect on the economy is a topic of
dispute.
What is clear, however, is that the declining costs of autonomous systems and a
corresponding decrease in manufacturing costs is propelling offshored manufacturing capacity
back to US soil.49 The automobile industry led the way in robotics adoption in an effort to drive
down labor costs.50 But the automobile industry cannot carry the weight of the manufacturing
sector in this adoption effort, much less spur the rest of the economy. In the next ten years, the
pharmaceutical, medical devices, electronics and food and beverage industries are all expected to
invest heavily in automated manufacturing infrastructure, but the forecasts for just how much are
imprecise and a matter of speculation.51 Industry optimists at Robotics Industry Association (RIA)
estimate that only 10% of US firms that could benefit from manufacturing automation have
adopted it so far—the implication being manufacturing automation represents a vast opportunity
for American firms to both boost productivity and reduce costs.52
The US economy has shifted since the heyday of manufacturing. American firms
increasingly provide services instead of physical products, and manufacturing now only comprises
12% of US GDP.53 Meanwhile, the types of reliable, precision systems required to suitably replace
humans in the services sector are still one-of-a-kind or hand-assembled units that remain cost
prohibitive. Compared to the manufacturing and industrial robot segments, service sector
applications—including both robotics and the application of sophisticated artificial intelligence to
replace human analysis—still represents a niche market in its infancy.54 By most accounts, the US
services sector—responsible for nearly 80% of nation’s GDP—will be unlikely to reap the benefits
from robotics-augmented productivity before the 2020 timeframe.55
The downside to an increasingly automated global economy is and will continue to be the
elimination of human jobs that accompanies the technological progress. While study data is sparse,
anecdotal evidence is mounting of an oft-repeated pattern where unskilled and semi-skilled labor
15
is replaced by technology.56 The prospect of losing thousands of jobs to increasingly automated
factories is likely to generate political repercussions designed to slow the inevitable shedding of
unskilled and semi-skilled labor positions. What is less clear is just how much the effects of the
recent US recession will dampen the fervor with which unions move to preserve status quo jobs
and by extension, erode the cost saving potential of autonomous systems. Again, hard data is
elusive, but anecdotal evidence indicates labor unions are increasingly willing to make concessions
in the hopes of saving at least some manufacturing jobs instead of losing them all to offshoring.57
On the other side, robotics proponents forecast aggregate job gains from the expected
increase in productivity and the accompanying growth in “downstream activities,” such as supply
chain activities, shipping, storage, and other supporting services.58 One study observed a set of 76
companies adopting industrial robotic technology, which resulted in the creation of 294,000 US
jobs over the course of a three year period.59 Robotics analysts currently estimate at least two
million aggregate jobs will be created worldwide by the year 2020 due to such activities.60 While
such an assessment sounds like cause for celebration, a bit of perspective is required. As the US
manufacturing sector alone lost two million jobs in a single year between 2008-2009, the aggregate
jobs forecast—if accurate—will not be the sole economic savior many are making it out to be.61
A secondary benefit may come from a renewed focus on engineering and production teaming,
which will strengthen the US historical core competency of product innovation, but the economics
of this shift are difficult to predict as well.62
While many factors make predicting the future effects on the US economy problematic,
one conclusion is certain. The potential productivity gains American industry—and later services
firms—could reap is not only open to American firms. Despite the onshoring movement and
increasing adoption of commercial robotics, the U.S. trails three other nations in robotic usage.
South Korea, Japan, and Germany all employ more robots per human worker than does the U.S.63
And the trend amongst US competitors is set to follow the path that killed the American
manufacturing industry in the first place: the world’s low-cost labor leaders, not content to lose
manufacturing dominance, are increasingly adopting robotic manufacturing techniques to continue
competing on cost. Proof? In 2013, China became the world’s leading importer of industrial robots,
purchasing one of every five produced worldwide.64
Even as American firms increasingly automate production, those same labor cost savings
will remain available to any firm—American or otherwise—willing and able to invest in the
technology. In other words, American firms will have to remain competitive through the use of
automated manufacturing, but its global use means robotic systems will become a minimum
requirement as opposed to a source of unique competitive advantage. This environment is hardly
a recipe for rescuing the US economy from its compounding national debt trap.
In order to ensure America’s industrial base is doing all it can to remain competitive on a
global scale, the US government needs to take proactive measures. First, the US government
should incentivize the adoption of robotic and autonomous systems in order to boost aggregate
worker productivity. Second, the U.S. must face head-on the pending shock to the jobs market
caused by robotics displacing human workers. While no single, easy solution exists, predicting the
types of jobs that will be displaced is straightforward. Thus, the government and industry should
team to make adjustments to the education and vocational training systems necessary to provide
the types of skills displaced workers will need to reenter their industries. It will not be an easy
transition for individuals or for the nation, but it is an entirely necessary one for the sake of the US
economy—and by extension, the future of America. (Lt Col Rob Masaitis, USAF)
16
ESSAY 2: ISRAELI LEADERSHIP IN ROBOTICS AND AUTONOMOUS SYSTEMS
Over the past decade, Israel’s successful integration of unmanned systems into its armed
forces helped the country become the planet’s leading exporter of unmanned air systems (UAS).
Israel’s success is firmly rooted in unique factors directly related to the country’s foundation,
historical experience and a hostile regional environment. These forces and the country’s small size
have endowed its people with a sense of urgency, common cause, and closeness—amplified by
the relationships forged through compulsory service in the IDF—that result in the pragmatic
approach to innovation that has yielded such outstanding results.
A young democracy founded after the trauma of the Holocaust and thrust immediately into
almost constant conflict with its Arab neighbors, Israel has felt the pressure of a constant existential
threat for much of its history. The challenge of creating a new state in a politically hostile
environment with scarce natural resources forced Israelis to find new, creative and innovative
solutions to the challenges they faced. From the outset, Israel has depended on and consistently
invested in the one strength that it had: its people. Succeeding waves of immigration brought
diversity and critical human capital that bolstered the nation’s considerable investment in
education, science, and technology and buttressed an entrepreneurial spirit unrivaled in the region.
This well-tended human capital enabled Israel to develop and maintain a qualitative edge that
allowed it to face numerically superior foes without exposing the Israel Defense Forces (IDF) to
the high casualties that would have devastated its small population.
Compulsory military service by all men and women over the age of eighteen is a shared
reality for most Israelis that gives the IDF tremendous influence in shaping the nation’s culture
and education. Soldiers acquire technical skills and a personal understanding of the military’s
needs before pursuing college degrees or entering the workforce. This experience is reinforced and
kept current by the continued obligation to serve in the IDF reserves. Compulsory service and the
prioritization of Israel’s security underpin a tight working relationship between the IDF and the
defense industry that enables innovation and rapid adoption of new defense products. Specialized
agencies and operational units collaborate with defense companies during all stages of
development and testing. The resulting shorter feedback loops and constant learning produce an
optimized development cycle for cutting edge technologies.
The Israeli acquisition system mirrors the pragmatism and collaboration evident in product
development. With persistent security challenges, rapid fielding of the latest systems is a shared
priority of the IDF and industry. The closeness of Israeli society allows field commanders to
request solutions to operational needs and to push hard for rapid acquisition. Sharing a high
tolerance for risk and improvisation characteristic of Israeli culture, commanders and developers
frequently collaborate on field-testing by deploying innovative products into operational use
before the end of the development cycle. Greater risk tolerance and a willingness to accept an
“80% solution” in order to get system into the soldiers’ hands dramatically speed the transition
time from prototype to program of record and accelerate the rate of adoption of new products.
For autonomous systems, the IDF has coupled the advantages detailed above with a
building block approach to development and testing. Unmanned system innovations in one domain
are leveraged to advance research in another. For example, a successfully tested and fielded control
algorithm is the natural starting point for the next unmanned platform. This “Lego” or building
block approach enables more rapid development and eases the test burden by relying on already
validated sub-systems. This methodology has allowed the IDF to acquire and upgrade unmanned
systems rapidly while promoting interoperability across its forces.
(BG Mordechay Baruch, Israeli Defense Force and Mr. David Mico, Department of State)
17
ESSAY 3: RAS INTEGRATION INTO LOGISTICS
While DoD logistics demands are complex and variable, projected and commercially-
proven RAS technology has the potential to deliver enormous labor and capital productivity
improvements. An analysis of common DoD logistics processes shows substantial overlap and
opportunity for integration of commercially available industrial robots. In short, it’s time for the
DoD logistics to embrace RAS.
Commercial trends in industrial robot development improve their utility for DoD logistics
application. Such commercial trends include: improved human-robot collaboration, greater robot
flexibility and user-friendliness, uncomplicated configuration, and low price points.65 State of the
art software graphical interfaces and intuitive reconfiguration and programming tools have
lowered user skill requirements and increased the flexibility of robotic applications. For example,
the Baxter robot simplifies programming by recording its physical movements as the user guides
the product’s arms and manipulators through the desired work steps.66
Many RAS technologies for DoD logistics application exist or are in development today:
Logistics Unmanned aerial vehicles (UAVs): The Marine Corps pioneered the use of
UAVs in tactical lift with its application of the remotely operated K-MAX helicopter in
Afghanistan.67 A more ambitious project by DARPA, the Areal Reconfigurable Embedded
System, aims to create a tactical resupply UAV.68 While the use of UAV technology in strategic
lift applications may not remove humans from cargo aircraft in the near future, the potential to
reduce flight crews below two is a soon-to-be-realized objective.
Autonomous Ground Vehicles (AGVs): AGVs offer obvious logistics utility for mail
delivery, material handling in depots and distribution centers and cargo handling. AGV designs
can operate under remote user control, in “follow-me” modes, or using a priori user-specified
route plans. For example, the Autonomous Mobility Appliqué System can transform an existing
cargo truck into an AGV to minimize risks to supply convoys.69 An AGV model to watch for DoD
strategic airlift is Frankfort International Airport’s implementation of pilot-controlled AGVs to
support heavy aircraft taxi from parking to point-of-takeoff to reap fuel cost savings.70
Warehouses: Commercial firms like Amazon and Diafuku already capitalize on AGVs to
dramatically improve productivity in factories, warehouses and distribution centers.71 The use of
autonomous material handling equipment also enables firms to capitalize on big data applications
through interfaces with warehouse, logistics and transportation management systems.72
Remote Presence Technology: When combined with precise manipulators, 3-D imaging,
and high bandwidth networks, remote presence technology allows high skilled maintenance
artisans to execute complex tasks on work pieces at remote distances. SRI International has
adapted this technology with light-weight, dual-arm actuators mounted on unmanned ground
robots to permit explosives experts to disarm bombs remotely.73 Such capability would allow the
DoD to affect depot-level repair in remote locations using artisans working in DoD facilities on
the other side of the globe.
Exoskeletons: Operational logistics personnel can reduce fatigue associated with repetitive
lift-and-carry tasks and amplify human strength through the use of exoskeletons. Lockheed Martin
has even developed logistics specific exoskeleton that lightens user loads and reduces fatigue with
an unpowered design.74
Adoption of robotics technology by the DoD logistics community will yield productivity
improvements. Over the long-term, such improvements will enable DoD to invest scarce budgetary
resources in warfighting technology and personnel thereby placing greater emphasis on the tooth
rather than on the tail. (Mr. Terence Emmert, Department of Defense)
18
ESSAY 4: OPPORTUNITIES FOR MANNED-UNMANNED TEAMING (MUM-T)
Today the U.S. faces growing, multi-axis threats and the need for presence in more places
than ever before. Given declining defense budgets, the U.S. must find an economical approach to
sustained conflict against global extremism and simultaneously prevent the overuse of high-end
capabilities designed to deter near-peer threats.75 Much like the previous offset strategy’s focus on
technology, the development of RAS may serve as a cornerstone of the next offset strategy. RAS
can provide both an economical means to address the low-end spectrum of military operations and
an effective means to mass against a high-end peer threat.76 To reap the full benefits of RAS, DoD
should employ the concept of Manned-Unmanned Teaming (MUM-T).
For MUM-T to cornerstone a truly effective strategy, the U.S. must develop Concepts of
Operations (CONOPS) in each domain that capture the true value RAS can provide the joint force.
In his book, Warfighting and Disruptive Technologies, Terry Pierce stated, “A study of modern
warfare suggests that whoever is first to combine new technologies with disruptive doctrine can
gain a decisive advantage. Conversely, a military that is slow to adapt new ways of fighting to
technological advance opens itself to catastrophic defeat.”77 Thus, the challenge remains: can DoD
describe the realm of the possible with RAS?
To meet this challenge, DoD should envision how MUM-T will benefit the force across
every discipline and in every regime. DoD should first analyze the portfolio of manned platforms
and then leverage RAS to increase effectiveness or to fill gaps in critical vulnerabilities. Instead of
perfecting a robotic system that can “do it all,” services should focus on a “system-of-systems”
approach that enables humans and robots to perform functions for which they are best designed.
Examples of such teaming approach exist on land, in the air, and in the sea:
Ground: Through a MUM-T focus, manpower intensive convoy operations can be
transitioned to a “pack” of manned platforms with unmanned follower ground vehicles.78 Israel
operationally employed this very concept in their latest conflict in Gaza. What used to take scores
of personnel could be accomplish with a mere handful, plus be exponentially safer in the face of
Improvised Explosive Device (IED) tactics.
Sea: The Office of Naval Research (ONR) is currently developing Unmanned Surface
Vehicles (USVs) that defend larger manned assets, such as destroyers and carriers.79 USV arsenal
ships, sailing in formation with their manned counterparts, would increase the overall lethality and
effectiveness of destroyers and provide a critical defensive edge against advanced enemy missile
systems. Likewise, future Unmanned Undersea Vehicles will be capable of autonomously tracking
enemy submarines and ships, creating “hunter-killer” teams between the manned and unmanned
subsurface force.
Air: Opportunities for MUM-T abound in supporting both our legacy aircraft and the
newer F-35 Joint Strike fighter (JSF). From unmanned refueling aircraft to Electronic Warfare,
teaming UAVs with manned aviation will improve effectiveness and efficiency while significantly
enhancing survivability through digital networking and shared sensor data.
Plentiful opportunities for effective teaming of manned and unmanned platforms exist. To
exploit the full potential of RAS, DoD must create opportunities for successful integration of
MUM-T into military operations and develop CONOPs that support and embrace RAS as a critical
enabler. Through the adoption of MUM-T, DoD can expand security and cooperative defense
initiatives around the world while preserving the US ability to provide suitable offset against near-
pear aggression. In short, MUM-T can enable RAS to serve as an effective and economical
component of the next off-set strategy. (LTC Kevin Murray, USMC)
19
ESSAY 5: DIB, GOOGLE, AND THE NEED TO STRENGTHEN THE DEFENSE
PRODUCTION ACT
US national security depends upon DoD’s ability to field innovation and cutting edge
technologies into our defense portfolio. With the decline of research dollars in the traditional
Defense Industrial Base (DIB) and the growth of powerful commercial companies like Google
overtaking new tech areas like autonomy, the US government must critically look at its approach
to engaging, leveraging and—at times—directing commercial markets. By updating our historical
and present day tools to meet tomorrow’s future challenges in developing and fielding robotics
and autonomous systems (RAS), the U.S. can be armed once again to bring the entire arsenal of
democracy to bear in furtherance of our national security objectives.
Historically, the U.S. has relied on a combination of patriotism and statutes to ensure access
to the manufacturing capability necessary to preserve national security. From Henry Ford’s
voluntary conversion of his factory capability to the war effort to the best physicists in the country
joining the Manhattan Project, the U.S. has effectively used patriotic calls to preserve national
security.80 Today, presidents enjoy a range of statutory tools to compel US companies to support
national security priorities.81 For example, Title I of Defense Production Act (DPA) allows the
president to impose priority contracts on domestic companies or individuals for goods and services
“necessary for national defense.”82 Thus, the government can trump other contracts through the
use of DPA prioritization, like it did when purchasing Mine Resistant Ambush Protected (MRAP)
vehicles for use in Afghanistan.83 Title III of DPA provides tools for the government to ascertain
the health and capability of a particular industry and take steps to ensure the country has the ability
to produce critical defense materials and goods.84
In his farewell address, President Eisenhower made the case for a vast, permanent
armament industry as to “no longer risk emergency improvisation of national defense.”85 While
America’s approach toward building a robust DIB has worked well historically, in recent years the
DIB has failed to keep pace with the commercial sector in RAS. In Ike’s day, the U.S. spent more
on military security than the income of all other US corporations combined.86 Today, Google’s net
worth is over twice the sum of the entire DIB; indeed Google could purchase any defense firm
simply with on-hand cash.87 Most troubling is how the commercial sector has outpaced the DIB in
terms of R&D. The combined R&D expenditures for the top five defense companies is less than
half of Google’s annual R&D.88 This disparity has grown more apparent with Google’s
procurement of top robotics firms. Within the last two years, Google has gobbled up eight of the
US top RAS firms.89 Google also possesses a seeming magnetic-pull on software engineering
talent. Both government research laboratory and industry officials lament the difficulty in retaining
talent in the face of Google job offers…especially in the autonomy area.90
What makes Google’s recent robotics purchases most troubling is that no one seems to
know what innovative breakthrough or robotics market the company is trying to pursue. Google’s
corporate values make clear that it seeks to provide “a great service to the world,” to “do things
that matter,” and above all “don’t do evil.”91 Recently, the chief of Google X—the main innovation
powerhouse within Google—reinforced the “don’t do evil” informal mantra and stated the
company’s desire to “actively make the world…a radically better place” even if that forsakes
opportunities for profit.92 While some are comforted by these grandiose visions of goodness, DoD
should be concerned if it fails to understand either Google’s innovation intentions or its
capabilities. In short, Google is not Ford and may not help the U.S. develop autonomous robotic
systems solely out of a sense of patriotic duty.93
Given the great number of unknowns about Google coupled with the declining R&D
20
investment by the DIB into RAS, the U.S. should take a serious look at a range of approaches—
from engagement to statutory compulsion—to ensure the nation can maintain its national security
edge in RAS. On the engagement front, the US government should appeal to a common set of
shared values with Google regarding autonomy. In the cybersecurity realm, Google has indicated
a willingness to collaborate with the government on “the defensive side of things.”94 In that same
vein, DoD should promote engagement through events such as DARPA challenges designed to
demonstrate RAS in natural and man-made disasters. Furthermore, DoD should recognize
Google’s leadership in autonomy and include Google in discussions about ethical and legal
implications of autonomous systems in warfare.
DoD should also encourage DIB companies to engage and partner with high-tech
commercial firms across a range of RAS projects. Even if a commercial firm like Google
voluntarily assisted DoD in furthering RAS technology during a national security crisis, they
would need DIB partners skilled in manufacturing, testing and deploying weapon systems in order
to bring systems to production. Through dialogue and joint partnerships in RAS, these firms may
find synergies in research endeavors—such as how to test autonomous systems cheaply and
effectively. While DoD has limited authority to compel DIB firms and pure commercial firms to
partner, DoD can facilitate interfaces and collaboration.
In addition to engagement, the president should assert his existing DPA authorities to
understand industry capabilities in RAS and to leverage Google personnel for national security
purposes. First, the president should direct an industry study under his Title III DPA authority to
understand the robotic industry capabilities and clarify which companies are capable of developing
autonomous systems for future defense use. In particular, this study should seek to understand how
the DIB can leverage innovations within the commercial robotics sector and determine whether
enough R&D dollars (between defense and civil companies) are being invested in autonomy
research. By identifying areas of overlap between the commercial and military sectors and
ascertaining gaps, DoD can better target limited R&D funds.
While Title III of the DPA can be applied to better our national security posture with respect
to RAS, Congress should also clarify Title I in the unfortunate event the president must compel
Google, or similar company, to assist in manufacturing RAS systems.95 It remains unclear whether
the government can force a company like Google, which does not make any goods, to produce a
product just because the firm possesses the know-how. While the statutory language implies that
the president can require performance under contracts for any entity he “finds capable” of such
performance, the statute expressly denies the president the ability to require purely employment
contracts.96 Thus, the question becomes whether the president could find a company legitimately
“capable” of producing a product when it currently sells no product nor accepts contracts for
production of goods.97 Enforcement becomes more problematic because the DPA is premised on
reassignment of contract “priorities” by jumping in the front of the production line.98 In Google’s
case, they do not currently accept production contracts for robotics nor service contracts to develop
software for other companies. In all, the ability for the government to exercise the Title I power of
the DPA toward Google remains unclear.
The U.S. should not play a passive role in understanding the deltas between commercial
and DIB capability. In light of lagging R&D investment in RAS by DIB and absorption of the best
robotics minds into commercial firms like Google, the US government should maximize
engagement opportunities and clarify applicability of the DPA to companies clearly capable of
RAS production, but who have yet to produce products or accept contracts. The U.S. can ill-afford
to start asking the tough questions about leveraging commercial RAS capabilities once on the
21
receiving end of an autonomous weapon strike. (Lt Col Linell Letendre, USAF)
ESSAY 6: RECOMMENDATIONS TO IMPROVE RAS TESTING
While RAS procurement poses unique challenges with requirements development and
lengthy acquisition processes, test and evaluation (T&E) may prove the most challenging aspect
to rapid RAS fielding. As one RAS program manager put it, “the testers may price autonomy out
of business.”99 To overcome this looming predicament, DoD must proactively develop and
resource a strategy that acquires the tools and technology necessary to test autonomous systems.
Why is T&E so challenging for RAS? First, contrary to normal T&E, testers are not
comparing the system’s actions against predicted actions but instead judging the decisions made
by the system.100 Evaluators must collect data to determine whether the system made a good choice
to accomplish a particular outcome and also understand why it made that decision and what its
degree of confidence was when making the decision.101 These testing difficulties are compounded
when placing a single autonomous system into an environment with multiple manned and
unmanned platforms. The resulting emergent behavior presents nearly an infinite amount of non-
deterministic responses to a given mission set.102 Additionally, when evaluating an autonomous
system during operational T&E, testers must assess the collaboration between the RAS and human
operator in order to determine if the system as a whole accomplishes the desired effects.103 Finally,
beyond just the testing of reliability and safety, T&E must yield trust in the warfighters, both those
operating RAS and those sharing the same battlespace.104 This last aspect—essentially testing RAS
for trust—greatly compounds testing difficulty.
While DoD has recognized these complexities, insufficient investment has been made and
numerous shortcomings exist in RAS testing.105 DoD has neither standardized testing framework
nor design of experiment methodology for RAS.106 DoD lacks a uniform modeling and simulation
(M&S) approach and does not possess a centralized database for comparing RAS performance.
Further, DoD has no standard way to judge a system’s level of autonomy nor has it adopted a
specific model for evaluating human-machine interaction or metrics for measuring human “trust”
in a system. Most troubling, however, is the lack of a consistent DoD strategy to enable the
development of the technology and tools necessary to validate autonomous systems.107
DoD should take a proactive, intentional approach to autonomous testing by:
Adopting a RAS T&E Strategy: Currently, DoD is developing RAS testing technology and
tools in an uncoordinated fashion. To correct this, DoD should assign an office of primary
responsibility, draft an overarching strategy for development of RAS T&E tools and technology
and then link resources to it. Within this strategy, DoD should establish target dates and
requirements for the development of: 1) standards for RAS testing frameworks, 2) a RAS M&S
approach, 3) metrics to evaluate human-machine interfaces, 4) measurements of human trust of
the unmanned system, and 5) a uniform method of determining levels of autonomy.
Creating a National Autonomous Testing Range: To propel such a strategy forward, DoD
should develop an autonomous test range with both robust M&S capability and physical test range
space. Such a test range would encourage standardization of testing frameworks and enable
centralized collection of data, which in turn would allow developers of autonomous systems to
learn and capitalize on unmanned platform testing from across multiple domains.
Almost three years have passed since the Defense Science Board laid out a series of
recommendations for the T&E community regarding autonomous systems, yet DoD has made
limited progress on developing the tools and techniques necessary to test RAS effectively.108 With
a unified RAS testing strategy and a national autonomy testing range, RAS can be rapidly
22
tested…and placed in the warfighter’s hands, where they belong. (Lt Col Linell Letendre, USAF)
23
Endnotes
1 Bob Work, “Deputy Secretary of Defense Speech – National Defense University
Convocation,” US Department of Defense, Press Operations, August 5, 2014, accessed April 1,
2015, http://www.defense.gov/Speeches/Speech.aspx?SpeechID=1873.
2 William Jennings Bryan, “William Jennings Bryan Quotes,” Good Reads, accessed
April 1, 2015, http://www.goodreads.com/quotes/41911-destiny-is-not-a-matter-of-chance-it-is-
a.
3 Trevor N. Dupuy, The Evolution of Weapons and Warfare (Fairfax: Da Capo Press,
1984), 1.
4 Paul Scharre, “How To Lose The Robotics Revolution,” War on the Rocks Blog, July
29, 2014, accessed 1 April, 2015, http://warontherocks.com /2014/07/how-to-lose-the-robotics-
revolution/.
5 Tom Andel, “American Robot Market Sees Double-Digit Growth in 2014,” Material
Handling & Logistics, October 30, 2014, accessed via ProQuest February 16, 2015; “World
Robotics Industrial Robots 2014,” International Federation of Robotics, accessed March 27,
2015, http://www.worldrobotics.org/uploads/media/Executive_Summary_WR_2014_02.pdf, 19.
6 Tom Green, “CES 2015: What Kinds of Robots Will People Actually Buy?” Robotics
Business Review, January 09, 2015, accessed March 29, 2015,
http://www.roboticsbusinessreview.com/article/ces_2015_what_kinds_of_robots_will_people_ac
tually_buy.
7 Terry Pierce, Warfighting and Disruptive Technologies: Disguising Innovation
(London: Frank Cass, 2004), 3-4.
8 Philip Kotler and Gary Armstrong, Principles of Marketing 15th Edition, Chapter 9,
Prentice Hall, 2013. 9 David P. Watson and David H. Scheidt, “Autonomous Systems,” Johns Hopkins APL
Technical Digest, 26.4 (2009).
10 Department of Defense, Unmanned Aircraft Systems Airspace Integration Plan, US
Task Force, Airspace Integration Integrated Product Team, Version 2.0, March, 2011.
11 Josh McBee, “Unmanned Aerial Vehicle (UAV) Manufacturing in the US,” IBISWorld
Industry Report, OD4424 (2012), 4.
12 Michael E. Porter, “The Five Competitive Forces that Shape Strategy,” Harvard
Business Review 86, no. 1, 78-86.
13 Josh McBee, “Unmanned Aerial Vehicle (UAV) Manufacturing in the US,” 15.
14 Ibid.
24
15 Roberto Buaron, et al., "Gaining advantage over competitors," McKinsey Quarterly,
(2000), 34-37.
16 Paul Scharre, “What To Look For In The DoD Budget,” War on the Rocks Blog,
January 28, 2015, accessed April 1, 2015, http://warontherocks.com/2015/01/what-to-look-for-
in-the-dod-budget/.
17 Dion Nissenbaum, “Drone Makers Take Aim at U.S. Market,” Wall Street Journal,
February 19, 2013, accessed April 1, 2015, http://www.wsj.com/articles/SB100014241278873
23764804578314501943739028.
18 Maksim Soshkin, “Drone industry hinges on defense spending and regulations,”
IBISWorld, Jan 2015, accessed April 1, 2015, http://media.ibisworld.com/ 2015/01/29/
droneindustry/.
19 Teal Group, “Teal Group Predicts Worldwide UAV Market Will Total $91 Billion in
Its 2014 UAV Market Profile and Forecast,” July 2014, accessed April 1, 2015
http://www.tealgroup.com/index.php/about-teal-group-corporation/press-releases/118-2014-uav-
press-release.
20 Valerie Insinna, “Drone Makers Hope to Corner Burgeoning Global Market,” National
Defense, May 2015, accessed 1 April, 2015, http://www.nationaldefensemagazine.org/
archive/2015/May/Pages/DroneMakersHopetoCornerBurgeoningGlobalMarket.aspx?PF=1.
21 Kelvin Chan, “The Big Story, Curbs shut US drone makers out of export markets,”
Daily Mail, accessed March 17, 2015, http://bigstory.ap.org/article/curbs-shut-us-drone-makers-
out-export-markets.
22 Terry Pierce, Warfighting and Disruptive Technologies: Disguising Innovation, 3-4.
23 Department of Defense, Unmanned Systems Integrated Roadmap FY2013-2038, 94.
24 Tom Barry, “Dysfunctional Drones Underscore Mission Mess at Homeland Security,”
Truth Out, 21 January, 2015, accessed April 1, 2015, http://www.truth-out.org/news/item/28645-
dysfunctional-drones-underscore-mission-mess-at-homeland-security.
25 Erik Schechter, “DoD seeks a common controller for many unmanned systems,”
Military Times, October 16, 2013, accessed April, 1 2015, http://archive.airforcetimes.com/
article/20131016/NEWS04/310160006/DoD-seeks-common-controller-many-unmanned-
systems.
26 Dave Mauumdar, “Requirements Debate Continues to Delay UCLASS RFP,” USNI
News, March 2014, accessed April 2, 2015, http://news.usni.org/2014/03/24/requirements-
debate-continues-delay-uclass-rfp.
27 Senior Vice President of industry, interview by Robotics and Autonomous Systems
Seminar. February 12, 2015
25
28 Dean Irvine, “Doing military’s dangerous, dull and dirty work,” CNN, February 16,
2012, accessed 1 April, 2015, http://www.cnn.com/2012/02/15/business/singapore-airshow-
drones.
29 “Our Work,” DARPA, accessed April 1, 2015, http://www.darpa.mil/our_work/.
30 “Twenty-five Teams From around the World to Participate in DARPA Robotics
Challenge Finals”, DARPA News, accessed April, 1, 2015, http://www.darpa.mil/NewsEvents/
Releases/2015/03/05.aspx.
31 Ibid.
32 Megan Garber, “Brain Drain Is Threatening the Future of U.S. Robotics”, Defense
One, June 30, 2014, accessed April 1, 2015, http://www.defenseone.com/technology/2014/06/
brain-drain-threatening-future-robotics/87562/.
33 John Markoff, “Google Adds to Its Menagerie of Robots,” New York Times, December
14, 2013, accessed April 1, 2015, http://www.nytimes.com/2013/12/14/technology/ google-adds-
to-its-menagerie-of-robots.html?_r=0.
34 Tina Nielson, “Business accelerators: a financial shot in the arm for startups,” The
Guardian, accessed March 20, 2014, http://www.theguardian.com/small-business-
network/2013/oct/24/business-accelerators-financial-startups.
35 Walter W. Powell, Kenneth W. Koput, and LaureSmith-Doerr, “Interorganizational
Collaboration and the Locus of Innovation: Networks of Learning in Biotechnology,”
Administrative Science Quarterly, 1996, accessed March 20, 2015, www.jstor.org/discover/
10.2307/2393988?sid=2116183826563&uiid=2460337935.
36 Terry Pierce, Warfighting and Disruptive Technologies: Disguising Innovation, 48.
37 Chuck Hagel, “Secretary of Defense Memo: The Defense Innovation Initiative,
November 15, 2014,” Defense Innovation Marketplace, accessed April 8, 2015,
http://www.defenseinnovationmarketplace.mil/resources/DefenseInnovationInitiative.pdf,
38 The Department of Defense, “The Long-Range Research and Development Program
Plan,” accessed April 8, 2015, http://www.defenseinnovationmarketplace.mil/LRRDPP.html.
39 Shawn Brimley et al., “Ideas to Action – Suggestions for the 25th Secretary of
Defense,” February 2015, accessed April 1, 2015, http://www.cnas.org/sites/default/files/
publications-pdf/CNAS_Ash-Carter_Briefing-Book_Feb2015_0.pdf.
40 Robotic Systems Joint Project Office, “Unmanned Ground Vehicle Interoperability
Profile Overarching Profile Version 0”, December 21, 2011, accessed March 31, 2015,
http://www.dtic.mil/ndia/2008groundrobot/Intlpanel1.pdf.
41 R. B. Normoyle, “Overview of the Joint Open Architecture Spectrum Infrastructure
(JOASI) Ontology for Spectrum Interoperability,” (Military Communications Conference, San
26
Diego, CA, November 18, 2013).
42 Israel Defense Forces (IDF) officials and engineers, interview by Robotics and
Autonomous Systems Seminar, Industry Studies International Travel, March 8-11, 2015.
43 Israeli robotics engineer, interviews by Robotics and Autonomous Systems Seminar,
March 11, 2015.
44 William Matthews, “Robot or Not? Army Dismisses Completely Soldierless
Battlefield,” ARMY, November 2014, accessed April 1, 2015, http://www.ausa.org/publications/
armymagazine/ archive/2014/Documents/11November14/ Matthew_November2014.pdf.
45 MarketLine, “Country Profile Series: United States In-depth PESTLE Insights”, July
2014, accessed February 11, 2015, http://store.marketline.com/Product/country_analysis_
report_united_states_in_depth_pestle_insights?productid=ML00002-032.
46 U.S. Congressional Budget Office, Updated Budget Projections: 2014 to 2024, April
2014, 6-9, accessed February 16, 2015, https://www.cbo.gov/sites/default/files/45229-
UpdatedBudgetProjections_2.pdf.
47 Zachary Roth and Daniel Gross, “President Obama Touts ‘Onshoring’: Is Made in
America Back?” The Daily Ticker, February 15, 2012, accessed February, 6, 2015,
http://finance.yahoo.com/blogs/daily-ticker/president-obama-touts-onshoring-made-america-
back-221759270.html.
48 Adams B. Nager and Robert D. Atkinson, “The Myth of America’s Manufacturing
Renaissance: The Real State of U.S. Manufacturing,” The Information Technology and
Innovation Foundation, January 2015, accessed February 13, 2015, http://www2.itif.org/2015-
myth-american-manufacturing-renaissance.pdf.
49 “International: New Robotics Tech Will Transform Work,” Oxford Analytica Daily
Brief Service, July 15, 2014, accessed February 16, 2015, http://www.relooney.com/
NS3040/000-Oxford-International_123.pdf.
50 Patrick Seitz, “Robots Not Job Killers, Trade Group Says As Sales Hit Record,”
Investor's Business Daily, August 4, 2014, accessed February 16, 2015,
http://news.investors.com/technology-click/080414-711685-north-american-robot-industry-off-
to-fastest-start-ever-in-first-half.htm.
51 Tanya M. Anandan, “Robots in 2014: Market Diversity, Cobots and Global
Investment,” January 10, 2014, accessed February 5, 2015, http://www.robotics.org/content-
detail.cfm/Industrial-Robotics-Industry-Insights/Robotics-in-2014-Market-Diversity-Cobots-
and-Global-Investment/content_id/4614.
52 “North American Robotics Industry Up 20% in 2012,” Robotics Industries Association,
October 29, 2012, accessed February 6, 2015, http://www.robotics.org/content-
detail.cfm/Industrial-Robotics-News/North-American-Robotics-Industry-Up-20-in-
2012/content_id/3752.
27
53 U.S. Department of Commerce, Bureau of Economic Analysis, BEA 14-55 News
Release: Widespread Growth Across Industries in Second Quarter 2014, November 13, 2014,
Table 5a, accessed February 16, 2015, http://www.bea.gov/newsreleases/industry/
gdpindustry/2014/pdf/gdpind214.pdf.
54 “International: New Robotics Tech Will Transform Work,” accessed April 1, 2015,
http://www.relooney.com/NS3040/000-Oxford-International_123.pdf.
55 Metra Martech, “Positive Impact of Industrial Robots on Employment,” International
Federation of Robotics, February 2013, 62-63, accessed February 16, 2015,
http://www.ifr.org/news/ifr-press-release/robots-to-create-more-than-a-million-jobs-by-2016-
295/. One notable exception is Rethink Robotics, a Boston-based firm that has developed a low-
cost, trainable system designed to work alongside humans—with a target price of $22,000 per
unit. Cecilia Kang, “New Robots in the Workplace: Job Creators or Job Terminators?”
Washington Post, March 7, 2013, accessed February 10, 2015, http://www.washingtonpost.com/
business/ technology/new-robots-in-the-workplace-job-creators-or-job-terminators/2013/03/06/
a80b8f34-746c-11e2-8f84-3e4b513b1a13_story.html.
56 Alana Semuels, “Factory Growth is No Job Machine,” Los Angeles Times, March 3,
2013, accessed February 6, 2015, http://articles.latimes.com/2013/mar/03/business/la-fi-
onshoring-walmart-20130303.
57 Ibid. 1-2.
58 One industry group pointed to a study in which 76 companies adopting industrial and
factory/warehouse robotic technology created a total of 294,000 jobs over the course of a three
year period. Colin Lewis, “Companies that use robots are creating jobs for people,” Robo Hub,
July 2, 2014, accessed March 29, 2015, http://robohub.org/more-jobs-in-companies-that-employ-
robots/.
59 Ibid.
60 Metra Martech, 10.; Mark T. Hoske, “Robots vs. Jobs: Jobs Win,” Plant Engineering,
January 2013, accessed February 10, 2015, http://www.plantengineering.com/ home/single-
article/robots-vs-jobs-jobs-win/2937982c1ad31bbdd34f0db2845d9501.html
61 Ibid., 4.
62 For example, see the improvements to GE’s product design and manufacturing effort
following its recent reshoring effort in Charles Fishman, “The Insourcing Boom,” The Atlantic,
December 2012, accessed February 6, 2015, http://www.theatlantic.com/ magazine/
archive/2012/12/the-insourcing-boom/309166/.
63 Simon Mundy, “South Korean Robots Lead the World,” FT.com, June 1, 2014,
accessed February 16, 2015, ProQuest online database.
64 Tanya Powley, “China Becomes Largest Buyer of Industrial Robots,” FT.com, June 1,
2014, accessed February 16, 2015, ProQuest online database.
28
65 International Federation of Robotics, accessed February 13, 2015,
http://www.ifr.org/industrial-robots/statistics/.
66 Rethink Robotics, “Baxter with Intera 3,” accessed March 19, 2015,
http://www.rethinkrobotics.com/baxter/.
67 Defense Industry Daily Staff, “Unmanned Hellos: The USMC’s Unmanned Cargo
Helicopters,” Defense Industry Daily, accessed March 18, 2015,
http://www.defenseindustrydaily.com/usmc-looks-for-an-unmanned-cargo-helicopter-06672.
68 DARPA, “ARES Aims to Provide More Front-Line Units with Mission-Tailored
VTOL Capabilities,” February 11, 2014, accessed March 19, 2015, http://www.darpa.mil/
newsevents/releases/2014/02/11.aspx.
69 Terry Brady, “U.S. Marines & Army Demonstrate Autonomous Mobility Applique
System,” Unmanned Systems Technology, August 28, 2014, accessed March 18, 2015,
http://www.unmannedsystemstechnology.com /2014/08/u-s-marines-army-demonstrate-
autonomous-mobility-applique-system/.
70 “Taxibot Concept,” Taxibot, accessed March 16, 2015, http://www.taxibot-
international.com/.
71 Supply Chain Digest Editorial Staff, “Supply Chain News: Amazon will not Make
Kiva Systems Available to Market for at Least Two Years,” Supply Chain Digest, accessed
March 20, 2015, http://www.scdigest.com/ontarget/14-03-31-1.php?cid=7944.
72 Amy Roach Partridge, “Materials Handling Meets Big Data,” Inbound Logistics,
accessed March 20, 2015, http://www.inboundlogistics.com/cms/article/materials-handling-
meets-big-data/.
73 “Products and Solutions,” SRI International, accessed March 20, 2015,
http://www.sri.com/engage/products-solutions/taurus-dexterous-robot.
74 “FORTIS Exoskeleton,” Lockheed Martin, accessed March 19, 2015,
http://www.lockheedmartin.com/us/products/exoskeleton/FORTIS.html.
75 Sam LaGrone, “Work: 2016 Budget Submission Reverses Five Year Defense Spending
Decline,” usni.org, January 29, 2015, accessed April 1, 2015,
http://news.usni.org/2015/01/29/work-2016-budget-submission-reverses-five-year-defense-
spending-decline.
76 Bob Work, “The Third U.S. Offset Strategy and Its Implications for Partners and
Allies,” US Department of Defense – Deputy Secretary of Defense Speech, January, 28, 2015,
accessed March 20, 2015, http://www.defense.gov/Speeches/Speech.aspx?SpeechID=1909.
77 Pierce, Warfighting and Disruptive Technologies Disguising Innovation, 48.
78 David Vergun, “Autonomous vehicles to exploit capabilities of machines, Soldiers,”
29
Army.mil, 30 June, 2014, accessed April 1, 2015, http://www.army.mil/article/129040/
Autonomous_vehicles_to_exploit_capabilities_of_machines__Soldiers/.
79 Staff Writer – Future Force, “Surface Autonomy is Heading for the Fleet,”
futureforce.navylive.mil, March 28, 2014, accessed April 1, 2015,
http://futureforce.navylive.dodlive.mil/2014/03/surface-autonomy-is-heading-for-the-fleet/.
80 Timothy J. O’Callaghan, Ford in Service of America: Mass Production for the
Military during the World Wars (Jefferson: McFarland & Co. Inc., 2009), 35-37; Jennet Conant,
109 East Place: Robert Oppenheimer and the Secret City of Los Alamos (New York: Simon &
Schuster, 2005), 54-55.
81 Interestingly, the Truman administration had considered and rejected a number of
statutory schemes that may have enabled the government to end the strike, including the
Selective Service Act of 1948, the Defense Production Act of 1950, and the Taft-Hartley Act.
Paulsen, “Youngstown Goes to War,” 227; McConnell, The Steel Seizure of 1952, 31-32.
82 Defense Production Act of 1950, US Code 50 Appx. (2014), § 2071.
83 Jared T. Brown and Daniel H. Else, “The Defense Production Act of 1950: History,
Authorities, and Reauthorization,” CRS Report R43118 (Washington, D.C.: Library of
Congress, Congressional Research Service, February 11, 2014), 9.
84 Defense Production Act of 1950, US Code 50 Appx. (2014), § 2093(a); Brown and
Else, 10.
85 Dwight D. Eisenhower, “Farewell Address to the Nation,” accessed February 19, 2015,
http://mcadams.posc.mu.edu/ike.htm.
86 Ibid.
87 Sandra I. Erwin, “Managing the Defense Industry: Stalinism or Smart Business?”
National Defense Magazine, November 2011, accessed February 19, 2015,
http://www.nationaldefensemagazine.org/archive/2011/November/Pages/ManagingtheDefense
IndustryStalinismorSmartBusiness.aspx; William J. Lynn III, “The End of the Military-Industrial
Complex,” Foreign Affairs 93, no. 6 (November/December 2014): 104, in EbsoHost (accessed
January 29, 2015).
88 Lynn, “The End of the Military-Industrial Complex”; Michael Casey and Robert
Hackett, “The 10 Biggest R&D Spenders World-wide,” Fortune, November 14, 2014, accessed
February 21, 2015, http://fortune.com/2014/11/17/top-10-research-development/ (reporting that
Google spent $8 billion in R&D in 2013 or 13.2% of its revenue).
89 Jacob Silverman, “Please, Don’t Be Evil,” New Republic, February 17, 2014, 2-13, in
EbsoHost (accessed 29 January, 2015).
90 Numerous defense research laboratory and industry officials, interviews with author,
various times over Eisenhower 2014-2015 school year.
30
91 Steven Levy, In the Plex: How Google Thinks, Works, & Shapes Our Lives (New
York: Simon & Schuster 2011), 150.
92 “Astro Teller Talks About Making Room for Failure,” Ben Johnson, Marketplace,
aired March 19, 2014, accessed March 20, 2015, http://www.marketplace.org/topics/tech/
marketplace-sxsw/astro-teller-talks-about-making-room-failure.
93 Google, in fact, already has a strained relationship with the U.S. government and has
resisted cooperation with the federal government through a variety of legal settings. For
example, Google has fought compliance with a DOJ subpoena to release child pornography data
and has challenged the gag order restricting Google’s release of numbers of FISA national
security requests. Ken Auletta, Googled: The End of the World as We Know It (New York:
Penguin Press, 2009), 191; Kashmir Hill, “Google Challenges Government Gag Order on
National Security Requests,” Forbes, June 18, 2013, 6, in EbsoHost (accessed 29 January, 2015).
94 Hill, “Google Challenges Government.”
95 As demonstrated by Youngstown, such compulsion lies outside the President’s
authority even in a time of war absent specific powers granted to the President by Congress.
Youngstown, 343 U.S. 759.
96 US Code 50 Appx. (2014), § 2071; Michael Littlejohn, “Using All the King's Horses
for Homeland Security: Implementing the Defense Production Act for Disaster Relief and
Critical Infrastructure Protection,” Public Contract Law Journal, no. 36 (Fall 2006): 1, 6, in
LexisNexis (accessed February 11, 2015), 5.
97 US Code 50 Appx. (2014), § 468(a); Littlejohn, “Implementing the Defense
Production Act,” note 27.
98 US Code 50 Appx. (2014), § 2071(a); Stuart B. Nibley, “Defense Production Act
Speeds Up Wartime Purchases,” National Defense, June 2006, 58, in ProQuest (accessed 6
February, 2015).
99 Defense official, Panel Discussion of 28 February, 2015, National Defense University
Autonomy Conference.
100 U.S. Department of Defense, Defense Science Board, Task Force Report: The Role of
Autonomy in DoD Systems, July 2012, 62, http://www.acq.osd.mil/dsb/
reports/AutonomyReport.pdf (accessed 20 March, 2015).
101 Ibid., 62-63.
102 Matthew Clark, “Test and Evaluation, Verification and Validation of Autonomous
Systems” (Safe and Secure Systems and Software Symposium, Air Force Research Laboratory,
Wright-Patterson AFB, OH, June 10, 2014), 3, accessed March 20, 2015, http://mys5.org/
Proceedings/2014/Day_1_S5_2014/2014-S5-Day1-02_Clark.pdf; Meltz and Gutterman,
“Verification of Safety for autonomous unmanned ground vehicles,” Electrical & Electronics
Engineers in Israel (IEEEI), 5 December, 2015, accessed April 4, 2015,
31
http://ieeexplore.ieee.org/xpl/articleDetails.jsp?reload=true&arnumber=7005895.
103 U.S. Department of Defense, Defense Science Board, Task Force Report: The Role
of Autonomy in DoD Systems, July 2012, 62, accessed March 20, 2015, http://www.acq.osd.mil/
dsb/reports/AutonomyReport.pdf.
104 Meltz and Gutterman, “Verification of Safety,” 1.
105 Ibid., 24.
106 Applying and institutionalizing the Design of Experiment (DOE) methodology to
OT&E has become a huge topic in the T&E community. V. Bram Lillard and Laura J. Freeman,
“Taking the Next Step: Improving the Science of Test in DoD T&E”, accessed March 20, 2015,
https://www.ida.org /~/media/Corporate/Files/Publications/IDA_Documents/OED/2014/D-
5101.pdf.
107 Matthew Clark, “Test and Evaluation, Verification and Validation of Autonomous
Systems,” 27.
108 Ibid., 12.